Mitosis-Mediated Intravasation in a Tissue-Engineered Tumor−Microvessel Platform

Published OnlineFirst September 18, 2017; DOI: 10.1158/0008-5472.CAN-16-3279 Cancer Integrated Systems and Technologies Research

Mitosis-Mediated Intravasation in a Tissue- Engineered Tumor–Microvessel Platform Andrew D. Wong1,2 and Peter C. Searson1,2,3

Abstract

Intravasation involves the migration of tumor cells across the cell interactions with functional microvessels and provide evi- local endothelium and escape into vessel ﬂow. Although tumor dence for a mitosis-mediated mechanism where tumor cells cell invasiveness has been correlated to increased intravasation, located along the vessel periphery are able to disrupt the vessel the details of transendothelial migration and detachment into endothelium through cell division and detach into circulation. circulation are still unclear. Here, we analyzed the intravasation of This model provides a framework for understanding the physical invasive human breast cancer cells within a tissue-engineered and biological parameters of the tumor microenvironment microvessel model of the tumor microenvironment. Using live- that mediate intravasation of tumor cells across an intact endo- cell ﬂuorescence microscopy, we captured 2,330 hours of tumor thelium. Cancer Res; 77(22); 1–9. 2017 AACR.

Introduction without the assistance of TAMs (5, 9, 10). These studies suggest that there are multiple pathways for intravasation, but the lack of Metastasis involves a sequence of steps including invasion, sufficient resolution has hampered our understanding of the intravasation, circulation, arrest, and extravasation that ultimately mechanism of tumor cell transendothelial migration (TEM) and result in tumor cell dormancy and/or growth at a secondary site detachment into circulation. (1). Intravasation, the entry of tumor cells into circulation, is a Recent advances in the development of in vitro microvessel critical step that occurs at both primary and secondary tumor sites models provide the tools to recreate the essential components of (2). The discovery of circulating tumor cells years after primary the tumor microenvironment and enable visualization of the treatment reveals that tumor cells continue to spread through the details of the metastatic cascade (11–13). Here we set out to vasculature and is a negative prognostic indication for cancer investigate the mechanism of intravasation of BCCs and to relapse (2, 3). Reducing intravasation may ultimately improve address the question, how do tumor cells cross endothelial junc- patient outcome; however, the details of these events are difficult tions to enter circulation? Using live-cell imaging in a tissue- to resolve, as it is a dynamic process that occurs at the interface engineered microvessel model of the tumor microenvironment, between tumor cells and the local tumor microvasculature. we analyzed over 2,330 hours of tumor cell interactions with Our understanding of the dynamics of intravasation is derived functional microvessels and found that intravasation events were largely from a relatively small number of intravital microscopy rare but predominately associated with mitosis. We quantified the (IVM) studies and few direct observations at the single cell level. deflection of peripheral tumor cells on the vessel endothelium For example, intravasation has been investigated in zebrafish and and provide evidence for a model where mitotic single-cell round- chick embryos (4, 5), but only directly observed in mouse models ing exerts a force on the endothelium that is sufficiently large to (6, 7). From these studies, we have learned that tumor associated transiently open endothelial cell–cell junctions and expose the macrophages (TAM) enhance the invasion of breast cancer cells tumor cells to shear flow, which pulls the daughter cells into (BCC) towards vessels through EGF and CSF1 chemotaxis (6, 8) circulation. To confirm that this is the dominant mechanism of and local expression of VEGF causes downregulation of endo- intravasation, we showed that tumor cells that extended protru- thelial cell–cell junctions, thus increasing the intravasation fre- sions across the interface did not intravasate. Similarly, tumor quency. However, tumor cells oversecreting growth factors are cells dividing in a larger perivascular space were unable to deflect also capable of intravasation across multiple in vivo systems the vessel endothelium and intravasate. These results demonstrate a simple, yet effective mechanism by which single tumor cells may undergo intravasation and provide a framework for understand- 1Institute for Nanobiotechnology, Johns Hopkins University, Baltimore, Mary- land. 2Department of Materials Science and Engineering, Johns Hopkins Uni- ing the physical and biological parameters that enable intravasa- versity, Baltimore, Maryland. 3Department of Oncology, Johns Hopkins Univer- tion through this pathway. sity School of Medicine, Baltimore, Maryland. Note: Supplementary data for this article are available at Cancer Research Materials and Methods Online (http://cancerres.aacrjournals.org/). Device fabrication Corresponding Author: Peter C. Searson, Johns Hopkins University, 100 Croft The tumor-microvessel platform was fabricated as described Hall, 3400 North Charles Street, Baltimore, MD 21218. Phone: 410-516-8774; previously (13). Briefly, high concentration rat tail collagen type I Fax: 410-516-5293; E-mail: [email protected] (Corning Inc.) is diluted to 7 mg/mL and neutralized with the doi: 10.1158/0008-5472.CAN-16-3279 manufacturer's recommended amounts of DI water, 10 PBS, 2017 American Association for Cancer Research. and 1 N sodium hydroxide. After neutralization, tumor cells are

www.aacrjournals.org OF1

Wong and Searson

introduced into the collagen solution to a final concentration of donkey anti-mouse IgG (H þ L) antibody conjugated to Alexa 5 105 cells/mL and injected around a cylindrical template rod Fluor 647 (A31571; Thermo Fisher Scientific) was incubated for (diameter 150 mm) within the polydimethyl siloxane (PDMS) 1 hour. Cells were washed for 10 minutes, three times between housing of the platform (Supplementary Fig. S1). After collagen fixing and incubation steps with PBS. gelation at room temperature, the rod is removed, leaving behind a cylindrical channel within the collagen gel. The channel is Image analysis fi m subsequently coated with bronectin (50 g/mL) to promote Intravasating tumor cells located at the ECM–vessel interface endothelial adhesion and spreading. Endothelial cells in along the middle of the vessel (i.e., vessel equator) were manually suspension are introduced into the channel at a concentration traced in ImageJ (NIH, Bethesda, MD) at each time-lapse frame. 6 of 5 10 cells/mL and allowed to settle and actively adhere to the Tumor cell circularity (i.e., shape factor, 4parea/perimeter2) was channel walls. After the endothelial cells have spread for about calculated as well as the maximum cell width perpendicular to the 2 hours, normal growth media (NGM) is perfused through the vessel wall. The latter was obtained by measuring the widths along < 2 vessel at a low applied shear stress ( 1 dyne/cm ) overnight. the entire length of tumor cell perpendicular to the vessel wall and fl Devices were typically con uent after 1 day and were switched to taking the maximum value with a custom written ImageJ plugin 2 higher shear stress ( 4 dyne/cm ) conditions for at least 24 hours (program provided in Supplementary Materials). These values before live-cell imaging. were plotted versus time. Cell motility descriptors (e.g., persistence, RMS speed, and directedness) were measured for tumor Cell lines and culture conditions cells located at the interface (n ¼ 38 cells continuously tracked for Human umbilical vein endothelial cells (HUVEC; Promocell), 12–17 hours from 4 independent microvessels) as well as in the human dermal microvascular endothelial cells (HMVEC; Lonza), bulk ECM (n ¼ 35 cells continuously tracked for 12–18 hours and VeraVec HUVEC-TURBOGFP (HVERA-GFP; Catalog. No. from 7 independent microvessels). Cell locations were approxi- HVERA-UMB-202100; Angiocrine Bioscience) were seeded in the mated in ImageJ from time-lapse images by manually tracing the cylindrical channel of the microvessel platform. Endothelial cells nucleus of each tumor cell every hour. Only data from tumor cells were grown in MCDB 131 (Caisson Labs) supplemented with tracked for a minimum of 12 hours were used, which reduced the 10% heat inactivated FBS (Sigma), 25 mg/mL endothelial mito- variance of each cell's average RMS speed, persistence, and direct- gen (BT-203, Biomedical Technologies), 2 U/mL herparin edness. Persistence is the shortest length connecting the start-to- m (Sigma), 1 g/mL hydrocortisone (Sigma), 0.2 mmol/L ascorbic end distance divided by the total path length acquired over the – – acid 2-phosphate (Sigma), and 1% penicillin streptomycin glu- entire duration of cell tracking. An average RMS speed for each tamine (Life Technologies). Dual-labeled MDA-MB-231 BCCs tumor cell was obtained by averaging the displacement over time (AntiCancer Inc.) were embedded within the collagen type I ECM for each one hour segment of a tumor cell's migration; for around the microvessel (14). Cancer cells were grown in RPMI example, a tumor cell tracked for 12 hours will exhibit 12 – (Corning Inc.) supplemented with 10% FBS and 1% penicillin individual RMS speeds that are combined as a batch average. streptomycin (Life Tech). All culture conditions were in humid- Directedness is the cosine of the cell's displacement angle, the fi i ed environments with 5% CO2 at 37 C. All cell lines were angle between the cell's start-to-finish vector acquired over the authenticated by their respective manufacturers and tested neg- entire duration of tracking with respect to either the direction of ative for mycoplasm. flow (i.e., along the length of vessel) or towards the vessel (i.e., direction perpendicular to the vessel wall). Live-cell imaging Fluorescence and phase-contrast images were obtained on a Nikon TE-2000 U microscope (Nikon Instruments Inc.). A 10 Frequency of intravasation, cell division, entry or exit events – objective was used for live-cell, time-lapse imaging and perme- from the ECM vessel interface ability experiments. Images were acquired every 10 minutes at Tumor cell intravasation, cell division, and entry/exit events are – equally spaced intervals along the entire length of vessel, typically rare (0 10 events per microvessel experiment) and individual 12 locations, 1 mm apart. Images were focused along the top/ tumor cells were tracked over varying amounts of time (e.g., – bottom (i.e., poles) and middle (i.e., equator) of each segment of 0.5 48 hours) because of cell migration into and out of the fi the vessel. imaging focal plane; due to these constraints, we quanti ed the frequency of these events over aggregate cell hours obtained over Immunofluorescence staining 26 microvessel experiments rather than as a fraction of cells CD133 (293C3, Miltenyi Biotec) mouse primary antibody was tracked. A tumor cell hour is the time spent by a single tumor incubated at 5 mg/mL per the manufacturer's recommended cell at the location of interest (e.g., interface or ECM) for 1 hour. fi – protocol. Mouse IgG2b isotype control (MA1-10427; Thermo For example, ve cells located at the ECM vessel interface tracked Fisher Scientific) and mouse anti-CD31 (37-0700; Thermo Fisher for 10 hours results in a total of 50 cell hours. Tumor cells at the – Scientific) antibodies were incubated at the same concentration ECM vessel interface or in the bulk ECM were tracked for a (i.e., 5 mg/mL). All primary antibodies were incubated on separate minimum of 30 minutes. All qualifying events (i.e., intravasation, samples. MDA-MB-231 and VeraVec were cultured on fibronectin cell division, entry, or exit from the interface) were summed and coated glass coverslips for 2 days prior to fixing and staining. Cells divided by the total number of tumor cell hours. were quickly washed with PBS and fixed with 3.7% formaldehyde for 3 minutes. Permeabilized samples were incubated in 0.1% Statistical analysis Triton X-100 for 3 minutes. Cells were blocked in 10% donkey Values are reported as mean SEM. The principle statistical test serum for 30 minutes. Primary and isotype control antibodies used was a Student t test (two-tailed assuming unequal variance). were incubated for 1 hour at room temperature. Secondary We considered P < 0.05 to be statistically significant.

OF2 Cancer Res; 77(22) November 15, 2017 Cancer Research

Mitosis-Mediated Intravasation in a 3D Microvessel

Results Tissue-engineered tumor–microvessel model To enable high-resolution imaging of the dynamics of intravasation into microvessels, human microvascular endothelial cells were cultured within 150-mm diameter cylindrical channels in a 3D collagen type I matrix. Dual-labeled MDA-MB-231 BCC (GFP histone tag, RFP cytoplasm) were embedded within the collagen matrix as either single cells or clusters (Supplementary Fig. S1). After a continuous endothelial monolayer is formed, usually within 24 hours after seeding, vessels were maintained under a constant flow rate of 1–2 mL/hour and time-lapse imaging was started the following day. The number of intravasating tumor cells has been positively correlated to the density of vessels 30 mmin diameter in tumor tissue (15), which suggests that vessels <30 mm do not contribute significantly to intravasation. Tumor vasculature is formed quickly, typically lacking smooth muscle cells and/or pericytes, and exhibits irregular architecture often with incomplete endothelial lining (16, 17); as a result, leaky tumor vessels will permit the heterogeneous extravasation of large par- ticles and red blood cells into the perivascular space and exhibit a global permeability significantly higher than normal vasculature Figure 1. (12, 17, 18). Our vessels exhibit permeability values ranging from Tumor cell invasion in tissue-engineered microvessels. A, Illustration of 3D 10 5 to 10 6 cm/s for BSA, which is comparable to that of leaky microvessel and the resulting 2D projected image from wide-field microscopy tumor vessels (12, 13). MDA-MB-231 is a human breast adeno- focused in the middle/equator of the vessel. B, Single dual-labeled MDA-MB-231 – carcinoma cell line from triple-negative breast cancer that has BCC extends multiple protrusions into the ECM vessel interface over 6 hours, which coalesce into a larger protrusion at the interface over 12 hours. C, Multiple been shown to readily metastasize to the brain in mouse models BCCs invading into the ECM–vessel interface using amoeboid motility, as shown in vitro (19, 20). MDA-MB-231 cells have been used in various by the nuclear deformation observed at 2.67 hours. Flow is from left to right. assays, such as transwell invasion studies (21–23), microfluidic adhesion and extravasation experiments (24–27), and IVM (20, 28). Insertion into the ECM–vessel interface BCCs that migrate to the microvessel insert protrusions into the Invasion interface between the ECM and the vessel endothelium (Fig. 1B). Our analysis is based on more than 2,330 hours of live-cell Tumor cells that successfully invade into the ECM–vessel interface imaging of BCCs across 26 individual microvessels (Fig. 1). Time- are typically within 50 to 100 mm of the vessel endothelium. lapse images were acquired every 10 minutes along the length of Extended processes that make contact with the vessel endotheli- the microvessel. When focused on the vessel equator, the endo- um are stabilized and exhibit flattened end-feet that progressively thelium is projected as a thin vessel wall perpendicular to tumor grow at the interface (Fig. 1B; Supplementary Video S2). The cells located in the ECM (Fig. 1A). Single BCCs in the matrix extended processes expand to a width of approximately 5 to generally exhibited a spindle-like morphology and often dis- 10 mm before the BCC is capable of squeezing into the interface played a rounded cell body with one or more protrusions at the (Fig. 1C). If the inserted BCC is part of a cluster, then additional leading edge while migrating, a characteristic of amoeboid migra- tumor cells may collectively migrate into the ECM–vessel interface tion (Fig. 1B; refs. 13, 29). Because the matrix is relatively dense, along etched tracks (Supplementary Video S1). Once inserted into migration involves proteolytic degradation of the collagen ECM, the interface, the cells typically exhibit a flattened egg shape at the which often creates etched tracks that are used by other cells to vessel poles (i.e., top/bottom) and appear elongated and spindle- facilitate rapid directed migration (Supplementary Video S1; refs. like in the projected equatorial view (i.e., middle/sides; Fig. 2A 29-31). Tumor cells in the ECM migrate at an average RMS speed and B). The fraction of tumor cells at the ECM–vessel interface of 0.20 0.036 mm/min with a persistence of 0.32 0.034 (n ¼ observed to migrate into the interface is 7.7% (17 of 221 interface 35 cells, seven microvessels, tracked for 12–18 hours). These BCC BCCs) over the 2,330 cell hours. The majority of BCCs at the speeds are comparable to tumor cell invasion in dense collagen interface remain with only 3.8% (8 out of 221 interface BCCs) gels (6–10 mg/mL) where proteolytic degradation is the limiting migrating back into the ECM. Interfacial tumor cells were capable factor (13, 29). Our persistence values are consistent with BCCs of invading into the ECM typically only if a predefined, etched observed on 2D surfaces in the absence of a chemotactic gradient track existed in the local ECM (Supplementary Video S3). The (32). Analysis of the angular orientation of BCC migration reveals higher rate of tumor cell insertion into the interface versus no significant directedness (0.028 0.026) towards or away returning to the ECM may be due to the lack of available etched from the vessel; therefore, approximately half of all tumor cells are tracks leading away from the interface; furthermore, tumor cells randomly migrating towards the vessel. Similar analysis of BCC are confined to the vessel interface rather than invading into the migration in the direction of flow, parallel to the vessel, shows no ECM, which requires proteolytic degradation and amoeboid significant directedness (0.057 0.047). These results indicate motility. that there is no apparent directional bias due to mechanotaxis (e. Tumor cells at the ECM–vessel interface also exhibited an g., interstitial flow or pressure gradients) or chemotaxis (e.g., average RMS speed of 0.24 0.019 mm/min with a persistence growth factor gradients from perfused media). of 0.36 0.034 (n ¼ 38 cells tracked for 12–17 hours from 4

www.aacrjournals.org Cancer Res; 77(22) November 15, 2017 OF3

Wong and Searson

Figure 2. Imaging tumor cell invasion and intravasation at the ECM–vessel interface. A, Illustration of a 3D microvessel and cross-section displaying the imaging focal planes at the vessel equator and pole. Below are representative images of a 3D projection and cross-section of a confocal z-stack of a HUVEC microvessel immunoﬂuorescently stained for VE-cadherin (green) and nuclei (blue). B, Illustration and representative images of tumor cells located at the vessel equator and pole, where they appear spindle-like and ﬂattened, respectively.

independent microvessel studies). The RMS speed and persistence events, both daughter cells escape into circulation; however, we of these tumor cells are not statistically different from their values have observed unique cases where one daughter cell intravasates in the ECM (P ¼ 0.34 and 0.42, respectively) despite being whereas the other returns to the interface (Supplementary Video confined in two dimensions. Tumor cells in the interface S8). In addition, it is possible for both daughter cells to return to exhibit no preferential directedness (0.065 0.044) with the interface after TEM, possibly due to insufficient shear stress respect to flow. required to detach the cells completely from the ECM substrate (Supplementary Video S9). Single-cell intravasation As described above, BCCs imaged at the equator of micro- Analysis of intravasation was restricted to single cells at the vessels and residing at the ECM–vessel interface appeared equator of the microvessels to clearly distinguish the position and elongated with a low circularity. Cell shape descriptors (e.g., profile of tumor cells with respect to the ECM substrate and circularity and maximum cell width) were quantifiable in 6 of endothelium in the X-Y focal plane of imaging (Fig. 2A). Despite the 8 intravasation events (Supplementary Fig. S4). The onset these restrictions, we observed a total of 2,330 cell hours from 221 of mitosis results in an increase in cell circularity, and the BCCs at the ECM–vessel interface across 26 separate microvessels. associated rounding imposes a deflection on the endothelium From this subpopulation of interfacial cells, we observed eight rather than on the stiffer ECM (Fig. 3D–I). If this deflection is intravasation events, thus measuring an intravasation rate of 3.4 large enough to disrupt the local endothelial cell–cell junctions, 10 3/hour (i.e., 8 events over 2,330 cell hours). All intravasation the tumor cell is exposed to the vessel lumen where the shear events displayed a characteristic rounding of the individual tumor stress may pull it into circulation. All intravasation events were cell, migration across the endothelium, and escape into circula- associated with deflections 20 mm, whereas cell division tion. In 4 of these 8 cases, intravasation was clearly associated with events with smaller deflections did not result in TEM or intra- mitosis (Supplementary Fig. S2; Supplementary Videos S4 and vasation (Fig. 3J). Intravasating cells, while rounding, exhibited S5), as identified by the visible organization of chromosomes an average maximum width of 30.3 5.0 mm(n ¼ 6), whereas and/or cell division prior to tumor cell detachment into flow. The dividing cells that did not intravasate exhibited a smaller remaining four cases were not definitively related to mitosis but average maximum width of 18.5 1.5 mm(n ¼ 15); these clearly involved cell rounding (Supplementary Fig. S2; Supple- values are not statistically different (P ¼ 0.067, two-tailed mentary Video S6). Tumor cells proliferate at a rate of 0.016/hour Student t test). During intravasation, tumor cell invasion along (i.e., 37 events over 2,330 cell hours) at the ECM–vessel interface, the ECM–vessel interface typically stopped during mitosis, and which is comparable to that in the bulk ECM of 0.014/hour (i.e., TEM was detected by the sudden translation of the tumor cell's 19 events over 1,301 cell hours in n ¼ 6 microvessels). In Fig. 3A center of mass towards the vessel lumen, about half the max- and B, two representative mitosis-mediated intravasation events imum width ( 15 mm), before commencement of cell rolling show BCCs at the ECM–vessel interface undergoing mitosis, and detachment (Fig. 3K). Our mitosis-mediated intravasation which leads to cell rounding, TEM, and escape into vessel flow. model is summarized in Fig. 3L. Supporting evidence for the Although intravasation events are often preceded by mitosis, physiologic relevance of mitosis-mediated intravasation comes we also observed 33 tumor cell division events at the ECM–vessel from the observations that: (i) tumor cell division potentiates interface that did not result in intravasation (Supplementary Fig. detachment through downregulation of adhesion proteins S3); cell shape descriptors could be analyzed for 15 of these and results in monoclonal clusters of circulating tumor cells events. In Fig. 3C, a representative nonintravasating cell division (33, 34), (ii) alternative attempts at TEM that involved extend- event shows a single BCC rounding at the interface, which deflects ing cell protrusions into vessel flow were not successful (4), (iii) the endothelium but does not result in TEM (Supplementary and no intravasation was associated with cell division in a large Video S7). Both resulting daughter cells resume a spindle-like perivascular space, but rather invasion and dispersal along the elongated shape at the interface. In all 8 detected intravasation ECM–vessel interface.

OF4 Cancer Res; 77(22) November 15, 2017 Cancer Research

Mitosis-Mediated Intravasation in a 3D Microvessel

Figure 3. Intravasation of interfacial tumor cells within the tumor–microvessel model. A and B, Representative timeseries of intravasating dual-labeled MDA-MB-231 BCCs. Notice the onset of mitosis and chromosomal organization during TEM across a VeraVec-GFP endothelium. Flow is from left to right. C, Representative timeseries of BCC division resulting in two daughter cells that remain at the interface. Flow is from left to right. D and E, Cell circularity and maximum width versus time for intravasating BCC featured in A. F and G, Cell circularity and maximum width versus time for intravasating BCC featured in B. H and I, Cell circularity and maximum width versus time for nonintravasating BCC featured in C. J, Maximum circularity and width of intravasating, n ¼ 6 (red), and nonintravasating, n ¼ 15 (blue), BCCs dividing at the ECM–vessel interface. K, X–Y location of a representative BCC invading along the ECM–vessel interface, followed by cell rounding and TEM. L, Illustration of invasion and mitosis-mediated intravasation, where a tumor cell invades into the ECM–vessel interface and cell division results in deﬂection of the endothelium, TEM, and tumor cell escape into vessel ﬂow.

Doxorubicin inhibits cell division and intravasation an intravenously administered chemotherapeutic that is used to To assess the role of cell division in intravasation of BCCs, treat a variety of cancer indications, such as breast and bladder microvessels were perfused with a mitotic inhibitor, doxorubicin, cancers. As a small molecule anthracycline that passively diffuses

www.aacrjournals.org Cancer Res; 77(22) November 15, 2017 OF5

Wong and Searson

across cell membranes, doxorubicin halts cell division by inhibit- Perivascular space reduces mitosis-mediated disruption of the ing topoisomerase II and intercalating with DNA (35). Here, we endothelium continuously introduced doxorubicin through functional In our model, intraluminal pressure is one of several factors microvessels at a pharmacologically relevant concentration of (e.g., cell size, ECM stiffness) that dictate the force applied by a 0.2 mmol/L in normal growth media (NGM) for 24 hours (36). dividing tumor cell on endothelial cell–cell junctions. Microves- After doxorubicin introduction for 8 hours, we measured the sels are maintained under positive intraluminal pressure through proliferation and intravasation rates of tumor cells imaged in the relatively high vessel flow rates (1–2 mL/h), thus resembling equatorial plane at the ECM–vessel interface. With a combined arterioles where the endothelium is in direct contact with the total of 617 cell hours from about 60 different tumor cells, ECM. We can mimic post-capillary venules by reducing the flow we would predict approximately 8 cell division events and rate, thus decreasing the intraluminal pressure and allowing the two intravasation events from previously measured division vessel endothelium to contract; this results in a gap or perivascular (0.016/hour) and intravasation (0.0034/hour) rates in NGM space between the vessel wall and ECM. The perivascular space is a alone; however, we observed no proliferation or intravasation common feature among post-capillary venules in the brain where events in NGM supplemented with doxorubicin, which confirms immune cells preferentially extravasate and tumor cells can that doxorubicin is a potent mitotic inhibitor and further supports achieve greater dispersal by migrating along the interface mitosis at the ECM–vessel interface as a potential mechanism for (39, 40). Under these conditions, dividing tumor cells within tumor cell intravasation. the perivascular space exert a smaller force on the overlying endothelium and are less likely to physically disrupt the endo- Tumor cell division enables detachment from clusters thelium during TEM (Supplementary Video S13). We have observed intravasation from BCCs that collectively To assess the time required for mitosis-mediated intravasation, invade into the interface and overrun the vessel endothelium, we determined the time period between the initiation of cell thus bypassing TEM. Direct exposure to vessel flow, tumor cell rounding, exhibited when the tumor cell circularity surpassed 0.6, overgrowth, and cell division all increase BCC susceptibility to and tumor cell escape into circulation. Based on this criterion, the shear forces and detachment (Supplementary Video S10). This intravasation time varied from 20 minutes to 9.3 hours (Supple- condition is similar to tumor vasculature that is incompletely mentary Fig. S4). This range is consistent with intravasation times lined by endothelial cells and/or partially lined by tumor cells, on the order of hours observed by IVM in mouse models (6), and which is estimated to comprise up to 4% of the vascular surface here, the duration likely reflects variation in the cohesive strength area in colon carcinoma (16), thought to account for roughly of endothelial cell-cell junctions and remaining tumor cell focal 106 tumor cells per gram of tumor tissue shed per day, and thus adhesions to the ECM. Since at least half of all intravasation events permits the entry of both single and clusters of tumor cells were directly preceded by mitosis, this implies that mitosis-medi- (15, 16, 37). Here, the majority of exposed tumor cells detach- ated intravasation is faster than other mechanisms, at least in our ing into flow are readily identified as undergoing cell division, model. and several having completed cytokinesis, appear as doublets, or two-cell clusters (Supplementary Video S11). Although both monoclonal and polyclonal clusters of circulating Discussion tumor cells have been detected in mice (33, 34), our mito- Using a tissue engineered microvessel model (13), we have sis-mediated intravasation model provides a plausible mecha- imaged and characterized invasion and intravasation of single nism by which both single and two cell monoclonal clusters MDA-MB-231 BCCs. The predominant mechanism that facilitat- enter circulation. ed TEM and detachment was tumor cell division at the ECM– vessel interface. Alternative methods for TEM, such as extending TEM without mitosis leads to blebbing protrusions across the endothelium into vessel flow, typically TEM exposes tumor cells to vessel flow, but does not enable resulted in shedding portions of the BCC and did not lead to detachment. We observe single tumor cells at the ECM–vessel intravasation. To confirm the important role of mitosis in phys- interface that become directly exposed to vessel flow, but do not ically disrupting endothelial junctions at the ECM–vessel inter- intravasate over 5 days (Supplementary Fig. S5), likely due to firm face, we observed that dividing BCCs in a perivascular space do adhesion to the ECM substrate. Tumor cell division induces TEM not deflect the endothelium or intravasate. Although our main by physically straining the overlying endothelium and reducing focus is on intravasation across functional microvessels, we also cell adhesion through the disassembly of focal adhesions (38). observe intravasation from BCCs directly exposed to vessel flow After being pulled into vessel flow, BCCs often leave behind after displacing the endothelium; under these conditions, tumor protrusions anchored to the ECM substrate, which further indi- cell mitosis leads directly to cell detachment due to increased cates the competing role adhesion plays with shear stress-induced exposure to vessel flow and reduced adhesion to the ECM and/or detachment (Supplementary Fig. S6). In the case where tumor nearby tumor cells. Since detaching tumor cells are in the process cells at the interface extend protrusions across the endothelium of dividing, a significant fraction are observed to have completed and into vessel flow, the intraluminal shear stress and constriction cytokinesis and will be circulating doublets, which is consistent from the surrounding endothelium typically results in shed por- with in vivo reports of monocolonal CTC clusters (33). tions or blebs from the tumor cells (Supplementary Video S12) During a total of 2,330 cell hours imaged at the ECM–vessel and does not lead to intravasation over multiple days. This interface, we observed eight intravasation events. All eight events observation is consistent with that in zebrafish inoculated with involved tumor cell rounding and four events were definitively MDA-MB-435 melanoma cells where blebs lost from tumor cell associated with mitosis (Supplementary Fig. S2). Cell rounding membranes extended into microvessels did not lead to full cell resulted in deflection of the overlying endothelium rather than the intravasation (4). stiffer ECM substrate. The stiffness and height of adherent cells are

OF6 Cancer Res; 77(22) November 15, 2017 Cancer Research

Mitosis-Mediated Intravasation in a 3D Microvessel

known to transiently increase during cytokinesis while focal sion and potentially intravasation (43); however, 5% (3 out of adhesions disassemble and integrin-mediated signaling is inacti- 61 interface BCCs) of interfacial BCCs were observed invading vated (38, 41). If the force on the endothelium is sufficient to into the ECM–vessel interface during 8–24 hours of doxorubicin disrupt cell–cell junctions, then the tumor cell body becomes perfusion, which is comparable to the 7.7% (17 of 221 interface exposed to vessel flow, which initiates cell rolling, loss of cell BCCs) under normal growth media conditions and suggests that adhesion to the ECM substrate, and tumor cell escape into the pharmacologically relevant doses of doxorubicin were not circulation (Fig. 3L). The observations that: (1) all intravasating affecting tumor cell invasion and EMT to a large degree during cells exhibited a maximum width 20 mm, and that (2) the the exposure. majority of dividing cells less than 20 mm in maximum width In addition, a larger tumor cell size will generate a greater (10 out of 15) failed to intravasate are consistent with a threshold deflection on the endothelium during mitosis, resulting in force necessary to disrupt endothelial cell-cell junctions (Fig. 3J). higher stress and possible separation of endothelial cell–cell From the number of intravasation events imaged at the vessel junctions. The tugging force between endothelial cell–cell junc- equator (8 events over of 2,330 cell hours), we obtain an intra- tions (e.g., adherens and tight junctions) has been measured vasation rate constant of 3.4 10 3/hour. Although we limit our at approximately 0.5–8nN/mm2 of junctional area, and the quantification to events occurring at the vessel equator, which maximum force between two cells begins to saturate above captures about 10% of the vessel's surface area, our intravasation 70 nN (44). Therefore, the separation force necessary for TEM rate is normalized to observed cell hours, and thus is applicable to between endothelial cells depends on both junctional area and all BCCs at the ECM–vessel interface. This rate is approximately integrity. Weakened endothelial junctions due to matrix 4-fold less than the intravasation frequency of 10 2/h reported for stiffening has been correlated to increased vessel permeability the MMTV-PyMT mouse model of mammary carcinoma using and leukocyte trafficking (45). Interestingly, junctional strength IVM where nearly all intravasation events occurred within close between endothelial cells increases over the time-scale proximity of TAM along the local microvasculature (6). TAMs of minutes to compensate for exogenously applied forces were shown to transiently downregulate endothelial barrier prop- (44); therefore, sudden deflection forces, such as tumor cell erties through the secretion of VEGF, which enabled the intrava- rounding/mitosis, may be more likely to break cell–cell junc- sation of single tumor cells (6). Tumor cells have also been shown tions at the ECM vessel interface. to utilize TAM-independent mechanisms of intravasation Similarly, the structure and function of the endothelium plays a through the overexpression of growth factors such as HBEGF that key role in regulating intravasation. The cohesive force of local down-regulate endothelial barrier properties (10). The lower cell–cell junctions must be overcome to enable TEM (46), which intravasation rate observed in this model may be attributed to has been studied in the TEM of leukocytes (39, 45). Factors that the lack of TAMs, which is consistent with the finding that the downregulate barrier function, either locally secreted by tumor intravasation rate is increased by the presence of macrophages in a cells or due to inflammation, are likely to promote intravasation separate in vitro model (42). The inverse of our intravasation rate (6). Although factors such as EGF and VEGF increase the concen- suggests that BCCs spend on average approximately 294 hours, or tration of circulating tumor cells in tumor-bearing mice (6, 7, 10), 12 days at the interface before intravasating across an intact vessel; it has not been established whether this is due solely to down- however, the average time for mitosis-mediated intravasation to regulation of vessel barrier function or from modulation of tumor occur is 4.37 1.5 h (n ¼ 6; Supplementary Fig. S4) and is cell phenotype. consistent with IVM observations (6). Accumulating evidence suggests that subpopulations of tumor Tumor cell phenotype can influence the intravasation rate in a cells, specifically cancer stem cells (CSC) and/or progenitor cells, number of ways. For example, tumor cell speed at the ECM–vessel exhibit a higher metastatic potential due to their increased capac- interface determines the sampling rate at which the tumor cell can ity for self-renewal, invasion, and survival (47, 48). Furthermore, find local disruptions or triple point junctions in the endothelium CSCs may participate in the formation of functional tumor lined where there is an increased probability of successful TEM. The vessels, termed vasculogenic mimicry, that promote neoplastic tumor cell's proliferation rate determines the frequency of cell growth and tumor progression (49). The mechanism by which rounding and hence the opportunities for TEM due to force CSCs intravasate is not well known; however, vasculogenic mim- generated on the endothelium. The rate of tumor cell proliferation icry precludes the need for TEM during intravasation, and the at the ECM–vessel interface is 0.016/hour (37 events over 2,330 detection of circulating CSCs in the blood of patients with cell hours), which is comparable to the tumor cell division rate colorectal cancer is a negative prognostic indicator (50). The within the ECM of 0.014 0.0047/hour (19 events over 1,301 cell positive expression of CD133, a cancer stem cell marker detected hours in n ¼ 6 microvessels). Roughly 10% of dividing cells at the in subpopulations of various tumor types (e.g., prostate, breast, ECM–vessel interface were able to complete intravasation (i.e., colorectal, etc.; refs. 50, 51), on MDA-MB-231 BCCs is inconsis- TEM and detachment) under normal growth media conditions. tently reported (52, 53). The MDA-MB-231 BCCs used in this Introducing doxorubicin, a potent mitotic inhibitor and com- study did not overtly express CD133 in any subpopulation of cells monly used chemotherapeutic for triple negative breast cancer, grown on fibronectin coated glass interrogated by immunofluo- abolished tumor cell division after 8 hours of exposure; no rescence staining (Supplementary Fig. S7). The lack of CD133 intravasation events were observed over the recorded 617 cell positive cells may be due to clonal variation, but does not hours acquired from 61 different tumor cells, which further discount the role of CSCs on intravasation. This tumor-micro- suggests a link between proliferation and single/doublet tumor vessel model may be uniquely capable of investigating the intra- cell intravasation. Drawbacks to using mitotic inhibitors to assess vasation potential of CSCs by substituting a CSC-rich population metastatic potential include secondary effects on tumor cell within the ECM. In our mitosis-mediated intravasation model, an function. For example, doxorubicin is implicated in disrupting asymmetric CSC division event at the interface would enable a cytoskeletal function, focal adhesions, and thus tumor cell inva- CSC population to be maintained at the interface while shedding

www.aacrjournals.org Cancer Res; 77(22) November 15, 2017 OF7

Wong and Searson

differentiated tumor cells (54); these released cells can be col- mediated mechanism of intravasation. Tumor cell division at the lected downstream and compared to those at the interface. ECM–vessel interface exerts a positive force on the vessel endo- In addition to tumor cell phenotype, our results demonstrate thelium where a larger tumor cell size and stiffer ECM substrate that the probability a dividing tumor cell will intravasate depends will increase the deflection of the vessel wall and stress on on specific properties of the tumor microenvironment, such as individual endothelial cell–cell junctions. Mitosis in this envi- matrix stiffness, vessel barrier function, intraluminal pressure, and ronment reduces a tumor cell's adhesion to the ECM while matrix adhesion. ECM stiffness increases the mechanical force a increasing its size perpendicular to the interface and susceptibility tumor cell is able to exert on the overlying endothelium when to vessel shear stress and detachment. Since we have shown that undergoing cell division. High stiffness in breast tumor tissue is a mitosis is correlated to intravasation, tumor cells at the ECM– negative prognostic indicator for vascular invasion (55). Stronger vessel interface with a higher proliferation rate will have an endothelial cell-cell junctions will improve the ability of the vessel increased likelihood of TEM and escape into vessel flow. Strategies endothelium to remain intact despite the underlying force exerted that target tumor cell proliferation, improve endothelial barrier by a dividing tumor cell. Furthermore, if the intraluminal pressure properties, and reduce intraluminal pressure and ECM matrix decreases, then the endothelium as a whole may compress stiffness will address mitosis-mediated TEM and detachment and inward and the force of the dividing tumor cell is minimized may ultimately decrease the rate of intravasation. (Supplementary Video S13). However, if a tumor cell has dis- rupted the endothelium and is exposed to intraluminal flow, its Disclosure of Potential Conflicts of Interest adhesion to the vessel basement membrane may oppose the shear No potential conflicts of interest were disclosed. stress and resist complete detachment and escape (Supplementary Video S9). Authors' Contributions If we consider the rate constant for intravasation (kin)asa Conception and design: A.D. Wong, P.C. Searson processes that requires an energy barrier (kT) to be surmounted, Development of methodology: A.D. Wong, P.C. Searson Acquisition of data (provided animals, acquired and managed patients, we can write kin ¼ A exp(EA/kT), where A is the attempt provided facilities, etc.): A.D. Wong frequency and E is the activation energy. The attempt frequency A Analysis and interpretation of data (e.g., statistical analysis, biostatistics, is simply the proliferation rate and the Boltzmann factor, exp computational analysis): A.D. Wong, P.C. Searson (EA/kT), represents the probability that a mitotic tumor cell is Writing, review, and/or revision of the manuscript: A.D. Wong, P.C. Searson able to exert sufficient force to overcome the endothelial cell–cell Administrative, technical, or material support (i.e., reporting or organizing junctions at that location. Therefore, by taking our intravasation data, constructing databases): P.C. Searson rate (3.4 10 3/hour) over our proliferation rate (0.014/hour), we obtain the probability that a dividing tumor cell at the interface Grant Support intravasates is approximately 0.24. The energy barrier for intra- This work was supported by NCI Provocative Questions (NIH vasation, kT, likely depends on a variety of factors mentioned R01CA170629). The costs of publication of this article were defrayed in part by the payment of above, such as the strength of endothelial cell–cell junctions, page charges. This article must therefore be hereby marked advertisement in stiffness of the ECM, and tumor cell width, which in principle, accordance with 18 U.S.C. Section 1734 solely to indicate this fact. could be deduced from additional physical measurements. In summary, the use of a tissue-engineered microvessel model Received December 1, 2016; revised July 9, 2017; accepted September 7, 2017; of the tumor microenvironment has revealed a novel mitosis- published OnlineFirst September 18, 2017.

References 1. Chambers AF, Groom AC, MacDonald IC. Dissemination and growth of 9. Deryugina EI, Quigley JP. Chick embryo chorioallantoic membrane model cancer cells in metastatic sites. Nat Rev Cancer 2002;2:563–72. systems to study and visualize human tumor cell metastasis. Histochem 2. Meng S, Tripathy D, Frenkel EP, Shete S, Naftalis EZ, Huth JF, et al. Cell Biol 2008;130:1119–30. Circulating tumor cells in patients with breast cancer dormancy. Clin 10. Zhou ZN, Sharma VP, Beaty BT, Roh-Johnson M, Peterson EA, Van Rooijen Cancer Res 2004;10:8152–62. N, et al. Autocrine HBEGF expression promotes breast cancer intravasation, 3. Hayes DF, Cristofanilli M, Budd GT, Ellis MJ, Stopeck A, Miller MC, et al. metastasis and macrophage-independent invasion in vivo. Oncogene Circulating tumor cells at each follow-up time point during therapy of 2014;33:3784–93. metastatic breast cancer patients predict progression-free and overall sur- 11. Katt ME, Placone AL, Wong AD, Xu Z, Searson PC. In vitro tumor models: vival. Clin Cancer Res 2006;12:4218–24. advantages, disadvantages, variables, and selecting the right platform. 4. Stoletov K, Montel V, Lester RD, Gonias SL, Klemke R. High-resolution Front Bioeng Biotechnol 2016;4:e12. imaging of the dynamic tumor cell-vascular interface in transparent zebra- 12. Bogorad MI, DeStefano J, Karlsson J, Wong AD, Gerecht S, Searson PC. fish. P Natl Acad Sci USA 2007;104:17406–11. Review: in vitro microvessel models. Lab Chip 2015;15:4242–55. 5. Zijlstra A, Mellor R, Panzarella G, Aimes RT, Hooper JD, Marchenko ND, 13. Wong AD, Searson PC. Live-cell imaging of invasion and intravasation in et al. A quantitative analysis of rate-limiting steps in the metastatic cascade an artificial microvessel platform. Cancer Res 2014;74:4937–45. using human-specific real-time polymerase chain reaction. Cancer Res 14. Hoffman RM. The multiple uses of fluorescent proteins to visualize 2002;62:7083–92. cancer in vivo. Nat Rev Cancer 2005;5:796–806. 6. Harney AS, Arwert EN, Entenberg D, Wang YR, Guo P, Qian BZ, et al. Real- 15. Liotta LA, Kleinerman J, Saidel GM. Quantitative relationships of intra- time imaging reveals local, transient vascular permeability, and tumor cell vascular tumor cells, tumor vessels, and pulmonary metastases following intravasation stimulated by TIE2(hi) macrophage-derived VEGFA. Cancer tumor implantation. Cancer Res 1974;34:997–1004. Discov 2015;5:932–43. 16. Chang YS, di Tomaso E, McDonald DM, Jones R, Jain RK, Munn LL. Mosaic 7. Wyckoff JB, Wang Y, Lin EY, Li JF, Goswami S, Stanley ER, et al. Direct blood vessels in tumors: frequency of cancer cells in contact with flowing visualization of macrophage-assisted tumor cell intravasation in mammary blood. Proc Natl Acad Sci U S A 2000;97:14608–13. tumors. Cancer Res 2007;67:2649–56. 17. Hashizume H, Baluk P, Morikawa S, McLean JW, Thurston G, Roberge S, 8. Roussos ET, Condeelis JS, Patsialou A. Chemotaxis in cancer. Nat Rev et al. Openings between defective endothelial cells explain tumor vessel Cancer 2011;11:573–87. leakiness. Am J Pathol 2000;156:1363–80.

OF8 Cancer Res; 77(22) November 15, 2017 Cancer Research

Mitosis-Mediated Intravasation in a 3D Microvessel

18. Yuan F, Dellian M, Fukumura D, Leunig M, Berk DA, Torchilin VP, et al. 37. Bockhorn M, Jain RK, Munn LL. Active versus passive mechanisms in Vascular permeability in a human tumor xenograft: molecular size depen- metastasis: do cancer cells crawl into vessels, or are they pushed? Lancet dence and cutoff size. Cancer Res 1995;55:3752–6. Oncol 2007;8:444–8. 19. Carbonell WS, Ansorge O, Sibson N, Muschel R. The vascular basement 38. Yamakita Y, Totsukawa G, Yamashiro S, Fry D, Zhang X, Hanks SK, et al. membrane as "soil" in brain metastasis. PLoS One 2009;4:e5857. Dissociation of FAK/p130(CAS)/c-Src complex during mitosis: role of 20. Bos PD, Zhang XH, Nadal C, Shu W, Gomis RR, Nguyen DX, et al. Genes that mitosis-specific serine phosphorylation of FAK. J Cell Biol 1999;144: mediate breast cancer metastasis to the brain. Nature 2009;459:1005–9. 315–24. 21. Lee TH, Avraham H, Lee SH, Avraham S. Vascular endothelial growth factor 39. Owens T, Bechmann I, Engelhardt B. Perivascular spaces and the two modulates neutrophil transendothelial migration via upregulation of stepstoneuroinflammation. J Neuropathol Exp Neurol 2008;67: interleukin-8 in human brain microvascular endothelial cells. J Biol Chem 1113–21. 2002;277:10445–51. 40. Burden-Gulley SM, Qutaish MQ, Sullivant KE, Lu H, Wang J, Craig SE, et al. 22. Liang M, Zhang P, Fu J. Up-regulation of LOX-1 expression by TNF-alpha Novel cryo-imaging of the glioma tumor microenvironment reveals migra- promotes trans-endothelial migration of MDA-MB-231 breast cancer cells. tion and dispersal pathways in vivid three-dimensional detail. Cancer Res Cancer Lett 2007;258:31–7. 2011;71:5932–40. 23. Wang L, Ling Y, Chen Y, Li CL, Feng F, You QD, et al. Flavonoid baicalein 41. Matzke R, Jacobson K, Radmacher M. Direct, high-resolution measurement suppresses adhesion, migration and invasion of MDA-MB-231 human of furrow stiffening during division of adherent cells. Nat Cell Biol breast cancer cells. Cancer Lett 2010;297:42–8. 2001;3:607–10. 24. Chen MB, Whisler JA, Jeon JS, Kamm RD. Mechanisms of tumor cell 42. Zervantonakis IK, Hughes-Alford SK, Charest JL, Condeelis JS, Gertler FB, extravasation in an in vitro microvascular network platform. Integrative Kamm RD. Three-dimensional microfluidic model for tumor cell intrava- Biol 2013;5:1262–71. sation and endothelial barrier function. Proc Natl Acad Sci U S A 2012;109: 25. Fraser SP, OzerlatGunduz I, Onkal R, Diss JK, Latchman DS, Djamgoz M. 13515–20. Estrogen and nongenomic upregulation of voltagegated Naþ channel 43. Millerot-Serrurot E, Guilbert M, Fourre N, Witkowski W, Said G, Van Gulick activity in MDAMB231 human breast cancer cells: Role in adhesion. J Cell L, et al. 3D collagen type I matrix inhibits the antimigratory effect of Physiol 2010;224:527–39. doxorubicin. Cancer Cell Int 2010;10:26. 26. Jeon JS, Zervantonakis IK, Chung S, Kamm RD, Charest JL. In vitro model of 44. Liu Z, Tan JL, Cohen DM, Yang MT, Sniadecki NJ, Ruiz SA, et al. Mechanical tumor cell extravasation. PLoS One 2013;8:e56910. tugging force regulates the size of cell-cell junctions. Proc Natl Acad Sci U S 27. Liang M, Zhang P, Fu J. Up-regulation of LOX-1 expression by A 2010;107:9944–9. TNF-a promotes trans-endothelial migration of MDA-MB-231 breast 45. Huynh J, Nishimura N, Rana K, Peloquin JM, Califano JP, Montague CR, cancer cells. Cancer Lett 2007;258:31–7. et al. Age-related intimal stiffening enhances endothelial permeability and 28. Stoletov K, Kato H, Zardouzian E, Kelber J, Yang J, Shattil S, et al. Visualizing leukocyte transmigration. Sci Transl Med 2011;3:112ra22. extravasation dynamics of metastatic tumor cells. J Cell Sci 2010;123: 46. Reymond N, d'Agua BB, Ridley AJ. Crossing the endothelial barrier during 2332–41. metastasis. Nat Rev Cancer 2013;13:858–70. 29. Wolf K, Wu YI, Liu Y, Geiger J, Tam E, Overall C, et al. Multi-step pericellular 47. Wicha MS, Liu S, Dontu G. Cancer stem cells: an old idea–a paradigm shift. proteolysis controls the transition from individual to collective cancer cell Cancer Res 2006;66:1883–90; discussion 95–6. invasion. Nat Cell Biol 2007;9:893–904. 48. Pang R, Law WL, Chu AC, Poon JT, Lam CS, Chow AK, et al. A subpop- 30. Gaggioli C, Hooper S, Hidalgo-Carcedo C, Grosse R, Marshall JF, Harring- ulation of CD26þ cancer stem cells with metastatic capacity in human ton K, et al. Fibroblast-led collective invasion of carcinoma cells with colorectal cancer. Cell Stem Cell 2010;6:603–15. differing roles for RhoGTPases in leading and following cells. Nat Cell Biol 49. Yao XH, Ping YF, Bian XW. Contribution of cancer stem cells to tumor 2007;9:1392–400. vasculogenic mimicry. Protein Cell 2011;2:266–72. 31. Sabeh F, Shimizu-Hirota R, Weiss SJ. Protease-dependent versus -indepen- 50. Iinuma H, Watanabe T, Mimori K, Adachi M, Hayashi N, Tamura J, et al. dent cancer cell invasion programs: three-dimensional amoeboid move- Clinical significance of circulating tumor cells, including cancer stem-like ment revisited. J Cell Biol 2009;185:11–9. cells, in peripheral blood for recurrence and prognosis in patients with 32. Desai SP, Bhatia SN, Toner M, Irimia D. Mitochondrial localization and the Dukes' stage B and C colorectal cancer. J Clin Oncol 2011;29:1547–55. persistent migration of epithelial cancer cells. Biophys J 2013;104: 51. Richardson GD, Robson CN, Lang SH, Neal DE, Maitland NJ, Collins AT. 2077–88. CD133, a novel marker for human prostatic epithelial stem cells. J Cell Sci 33. Cheung KJ, Padmanaban V, Silvestri V, Schipper K, Cohen JD, Fairchild AN, 2004;117:3539–45. et al. Polyclonal breast cancer metastases arise from collective dissemina- 52. Liu TJ, Sun BC, Zhao XL, Zhao XM, Sun T, Gu Q, et al. CD133þ cells with tion of keratin 14-expressing tumor cell clusters. Proc Natl Acad Sci U S A cancer stem cell characteristics associates with vasculogenic mimicry in 2016;113:E854–63. triple-negative breast cancer. Oncogene 2013;32:544–53. 34. Aceto N, Bardia A, Miyamoto DT, Donaldson MC, Wittner BS, Spencer JA, 53. Croker AK, Goodale D, Chu J, Postenka C, Hedley BD, Hess DA, et al. High et al. Circulating tumor cell clusters are oligoclonal precursors of breast aldehyde dehydrogenase and expression of cancer stem cell markers selects cancer metastasis. Cell 2014;158:1110–22. for breast cancer cells with enhanced malignant and metastatic ability. J 35. Tacar O, Sriamornsak P, Dass CR. Doxorubicin: an update on anticancer Cell Mol Med 2009;13:2236–52. molecular action, toxicity and novel drug delivery systems. J Pharm 54. Morrison SJ, Kimble J. Asymmetric and symmetric stem-cell divisions in Pharmacol 2013;65:157–70. development and cancer. Nature 2006;441:1068–74. 36. Gabizon A, Shmeeda H, Barenholz Y. Pharmacokinetics of pegylated 55. Evans A, Whelehan P, Thomson K, McLean D, Brauer K, Purdie C, et al. liposomal doxorubicin: review of animal and human studies. Clin Phar- Invasive breast cancer: relationship between shear-wave elastographic macokinet 2003;42:419–36. findings and histologic prognostic factors. Radiology 2012;263:673–7.

www.aacrjournals.org Cancer Res; 77(22) November 15, 2017 OF9

Mitosis-Mediated Intravasation in a Tissue-Engineered Tumor−Microvessel Platform

Andrew D. Wong and Peter C. Searson

Cancer Res Published OnlineFirst September 18, 2017.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-16-3279

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2017/09/16/0008-5472.CAN-16-3279.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2017/11/02/0008-5472.CAN-16-3279. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.